债务融资、两权分离度与终极股东掏空及支持行为：基于面板门限模型的实证分析

Debt financing, control-ownership divergence and the ultimate shareholders ' tunneling and propping behavior: Empirical analysis based on panel threshold model

利用375家非金融类上市公司2007~2013年面板数据,分别以总债务、银行债务、商业信用、短期债务和长期债务为门限变量,构建面板门限回归模型,实证分析了不同债务状态下两权分离度对终极控制股东掏空与支持行为的非线性影响。结果表明:两权分离程度对终极控制股东行为的影响效应存在基于公司总债务、银行债务和短期债务等变量的门限特征,基于商业信用和长期债务的门限特征则不稳定。具体而言,当总债务比率或短期债务比率分别高于0.901及0.848时,两权分离度促使终极控制股东实施支持行为;当银行债务比率高于0.149时,两权分离度引起终极控制股东的掏空行为。本文的研究结论为中小股东和外部投资者对上市公司投资时评估该公司股权结构的合理性提供了一种参考依据。

Pyramid control structure is a common form of ownership structure in the world, its main feature is the separation between control rights and cash flow rights. This new type of control-ownership divergence usually leads to the tunneling behavior of ultimate controlling shareholder, while the complexity of the pyramid control structure provides a great convenience for the behavior. However, the control-ownership divergence will also promote the propping behavior of ultimate controlling shareholder in some conditions. Tunneling and propping are two kinds of opposite behaviors and they appear in different state of companies. In which state will the control-ownership divergence cause the ultimate controlling shareholder＇s tunneling behavior？ In which state will the control-ownership divergence lead to the ultimate controlling shareholder＇s propping behavior？ The relevant research is scarce. As debt financing is an important way of external financing, and it is also an important part of corporate governance. This paper attempts to study the nonlinear effect of the control-ownership divergence on the ultimate controlling shareholder＇s tunneling and propping behavior under different debt financing states. The goal is to provide a reference for dealing with the problem of corporate governance. Under the pyramid control structure, there are two different views on the influence of debt financing on the ultimate controlling shareholders＇ behaviors. One view is that debt financing may promote the tunneling behavior of ultimate controlling shareholders. The other view is that different types of debt have different effect on the ultimate shareholder＇s behaviors. There are three reasons for this phenomenon. Firstly, the variable used as a proxy for tunneling or propping is different. Some scholars use the ultimate controlling shareholders＇ control-ownership divergence as the indirect variable proxy for the tunneling behavior, while other scholars use the occupation of listing corporation funds as the variable proxy for the tunneling behavior. Secondly, some scholars only consider the ultimate controlling shareholders＇ tunneling behavior of the listed companies without considering the propping behavior. Thirdly, scholars only consider the linear relationship between the variables in the study, and they don＇t consider the nonlinear relationship of the interval effect. Taking into account the problems or deficiencies in the study, this paper does the following works. First of all, this paper finds a proxy variable that can simultaneously reflect the tunneling and propping behaviors, namely ＂the tunneling degree of the ultimate controlling shareholders＂. The positive value means tunneling behavior and the negative value means propping behavior. It then establishes panel threshold regression models, which well characterize the nonlinear relationship of interval effect between variables. The study further explores and highlights the specific interval of tunneling behavior or propping behavior. Finally, an empiricalanalysis is conducted to verify and clarify the conclusions of the study. This paper uses the panel data based on 375 non-financial listing corporations in China from 2007 to 2013, and constructs a panel threshold regression model. It uses total debt, bank debt, commercial credit, short-term debt and long-term debt as threshold variables, and anglicize the nonlinear effects of the control-ownership divergence on the propping and tunneling behavior of ultimate controlling shareholder under different debt states. In this study, the paper finds that the variables of total debt, bank debt and short-term debt have threshold effect, but the threshold effect of commercial credit and long-term debt are not stable. Firstly, when the total debt ratio or short-term debt ratio are higher than 0.901 and 0.848 respectively, the control-ownership divergence can promote the propping behavior of ultimate controlling shareholder, that is, the total debt and short-term debt highlights the governance effect. When the total debt ratio is lower than 0.901 or short-term debt ratio is lower than 0.848, the control-ownership divergence can cause the tunneling behavior, but the influence is not significant. Secondly, when bank debt ratio is higher than 0.149, the control-ownership divergence can cause the tunneling behavior. When bank debt ratio is lower than 0.149, the control-ownership divergence can promote the propping behavior, but the influence is not significant. The results of this paper provide a reference for small and medium shareholders and outside investors to evaluate the rationality of the company＇s ownership structure.